Chronic Kidney disease (CKD) is a long-term condition where kidneys don't work as well as they should.
It is associated with a decrease in kidney function over time and is a growing public health concern. It frequently occurs with a wide range of other conditions such as high blood pressure, heart disease, and diabetes. Glomerular filtration rate (GFR) is a test that is used to check how well kidneys are functioning by measuring the amount of blood that passes thought kidneys each minute and is used to assess the CKD stage. Patients with lower GFR are more likely to have adverse events (eg. death, kidney failure). Early treatment can prevent or delay CKD from becoming worse which often leads to more complications. The early stage of CKD does not cause any symptoms and only a few patients develop end-stage kidney disease. For new treatment development, it is important to identify patients that have a progressive decline in kidney function. Our proposed study aims to assess disease progression in CKD patients and evaluate how patient-related characteristics can predict future CKD decline and other adverse events. This will facilitate the definition of inclusion criteria that will improve patient recruitment in clinical studies.
Objective
Whilst only a small proportion of CKD patients progress to end stage renal disease, there is dearth of knowledge on factors predictive of CKD progression. Our proposed study aims to assess projection trajectories in CKD patients and evaluate how prior GFR decline and clinical characteristics can predict future CKD decline and other adverse events (ESRD, CVD and mortality). This will facilitate the definition of inclusion criteria for clinical trials, thus improving patient recruitment and trial success.
Method and Analysis plan
We will identify individuals with first ever recorded diagnosis of CKD between 2004 and 2018 in CPRD-HES linked data. CKD will be defined using two consecutive measurements of estimated glomerular filtration rate (eGFR) at a value of less than or equal to 75 ml/min/1.73m2, at least 3 months apart. CKD progression will be assessed using eGFR readings after CKD diagnosis and based on their eGFR slopes, patients will be categorised as slow or rapid progressors. By exploring a carefully selected set of modelling techniques (simple directed approaches, Bayesian clustering, Neural networks; please see justification below), we will assess whether rapid CKD progression can be predicted based on patient characteristics and comorbidities. Finally, we will use Cox regression model to compare risk of adverse outcome (mortality, end stage renal disease (ESRD)) among slow and rapid progressors.
The overall aim of this study is to evaluate and predict disease progression in patients with CKD.
Specific study objective
1. Assess progression of CKD patients with late CKD stage 2, 3 and 4 to determine an appropriate definition of a fast-progressor versus a slow/non-progressor.
2. Identify clinical and demographic characteristics of patients who are faster and more severe progressors and use statistical and machine learning techniques to predict the risk of rapid CKD progression based on patientsÂ’ characteristics and comorbidities.
3. Examine the association between the slope of eGFR over time and adverse events (ESRD, CVD and all-cause mortality).
Rationale
Our proposed study aims to assess projection trajectories in CKD patients and how prior GFR decline can predict future CKD decline and other adverse events (ESRD, CVD and mortality) in these patients. This will facilitate the definition of inclusion criteria that will optimise patient recruitment in clinical trials and thereby ensuring an adequate therapeutic window for patient enrolment into clinical trials. We will also explore the use of both statistical and machine learning approaches to examine which methods are better suited for this challenge
Lutz Jermutus - Chief Investigator - AstraZeneca Ltd - UK Headquarters
Alyshah Abdul Sultan - Corresponding Applicant - AstraZeneca Ltd - UK Headquarters
Domingo Salazar - Collaborator - AstraZeneca Ltd - UK Headquarters
Glen James - Collaborator - AstraZeneca Ltd - UK Headquarters
Irena Brooks-Smith - Collaborator - AstraZeneca Ltd - UK Headquarters
Jerry Wu - Collaborator - AstraZeneca Ltd - UK Headquarters
Jolyon Faria - Collaborator - AstraZeneca Ltd - UK Headquarters
Michail Doulis - Collaborator - AstraZeneca Ltd - UK Headquarters
Xiang Ji - Collaborator - AstraZeneca Ltd - UK Headquarters
HES Accident and Emergency;HES Admitted Patient Care;ONS Death Registration Data;Practice Level Index of Multiple Deprivation